Syncing LED's to Music

[Castro]

Hello Everyone,

I first want to say that this forum is awesome, I stumbled upon it through a google search and there is such a wealth of information here for people interested in electronics.

Okay, now that that is out of the way, on to my problem...I'm looking to create a circuit that takes as its input, left and right speaker signals, amplifies that signal, then separates it according to a range of frequencies by using a LPF, HPF and BPF. The output signal would then trigger high power LED's (3.9V 700mA)

Here is a circuit that I found on the internetz **broken link removed**

For each op amp, I wanted a gain of 10 so I'm planning to use 100K for each feedback resistor and 10K for each input resistor.

Concerning the active filters, I decided to go with a cutoff frequency of 100Hz on the low end, 12kHz on the high end and between 100Hz and 3000Hz for midrange.

Am i on the right route to do what im planning to do. Can someone explain how the darlington transistor play in all of this. I know its to amplify the current as much as I can to the LED's but I cant really grasp it mathematically and conceptually yet. Do I need to take into account the input voltage range of the speaker signals (change in volume)?

Sorry for the long thread, but I really want to complete this. Thanks!

 

[aljamri]

Simple LED needs no more than 10 mA.

 

[Castro]

im looking to use luxeon LED's with forward voltage at 3.7 V and rated up to 1400 mA

 

[audioguru]

The site with the schematic is asleep or dead. It is too bad you didn't copy the schematic and attach it here to your thread. Now nobody knows what you are talking about.

Your choice of the frequencies for your filters leaves ot a lot of sound from 3kHz to 12kHz.

 

[Castro]

Sorry about the dead link guys...

**broken link removed**


I've thought over the frequencies,
LP cutoff: 1kHZ
HP cutoff: 6kHZ
BP center: 3kHZ

Also, how much different would this be to implement if I use an electret microphone as pass the input signal? Is it as accurate as passing the signals directly?

 

[audioguru]

I see extremely simple 6dB/octave filters. maybe you should use active second-order filters.
The filters have such a small slope that the bandpass filter will reduce the level of the frequencies you want.

I have a Sound Level Indicator project installed in my family room. Its input is an electret mic and it picks up a pin dropped on the floor of the next room and loud stereo, TV and talking.

Your LEDs will blow up because nothing limits their current.

 

[Castro]

Thats what I'm not too sure about, how can I measure the max current at the output? And once I figure this out, would I just use an appropriate resistor value to bring the current down to just below the maximum allowed current of the load.

Since this is my first stab at this, I want to keep complications to a minimum, so a first order circuit would be fine for my objectives.

BTW, thanks audioguru, your posts help a lot.

 

[audioguru]

Look on the datasheet to see the minimum, typical and max allowed current for the LEDs.
Look at its minimum, typical and max forward voltage at the current you want.

Look at the datasheet for the darlington transistor to see its typical and max saturation voltage loss at the current you will use.

Then subtract the voltages from the supply voltage to detrermine the voltage that will be across the current-limiting resistor. Use Ohm's Law to simply calculate the resistor and calculate its power dissipation (heat).

 

[Castro]

^thanks audioguru, that makes sense...

OK, please pardon the super noob question, but can I power this circuit from a single 12V source if the LED's are rated at 3.9V, transistors @ 1.4V min, and op-amps @ 12v?

 

[audioguru]

can I power this circuit from a single 12V source if the LED's are rated at 3.9V, transistors @ 1.4V min, and op-amps @ 12v?

Of course you can if you calculate the current-limiting resistors correctly. These resistors will get hot so use big ones.

 

[Castro]

Hello Again...Im trying to simulate my filters using pSpice, but I'm having trouble with my analysis. Is it necessary to have a AC input when dealing with filters?

 

[audioguru]

Your filters are simple passive RC ones with a gradual slope of only 6dB/octave. Active filters with steeper slopes would be much better.
Your frequencies are too close together for such a gradual slope.
You can measure the outputs of the filters if you temporarily remove the darlington transistors.

 

[Dean Huster]

It seems like "color organs" used to be bi-monthly projects back in the day when there were a lot of different electronics hobbyist magazine titles out there. Find one of those and use the basic "front ends" and replace the usual incandescent "output" with LED drivers.

 

[Castro]

Your filters are simple passive RC ones with a gradual slope of only 6dB/octave. Active filters with steeper slopes would be much better.
Your frequencies are too close together for such a gradual slope.
You can measure the outputs of the filters if you temporarily remove the darlington transistors.

Eventually I want to implement butterworth filters into my circuit, but right now I cant even run a simple pSpice simulation lol.

Here's how I have it set up;

DC Voltage source @ 12V in series with a 2K resistor in series with a 80nF capacitor. These numbers should give a cutoff of 1kHZ. How can I simulate the frequency response using pSpice? DC sweep?

 

[canadaelk]

Dean:
You are right! This is so 70s
1870, Tesla would have been proud!
E

 

[audioguru]

If the 1k resistor is driven by a very low impedance generator, then it has a cutoff at 1kHz when the capacitor to ground is 80nF and the load impedance is high.
1kHz= -3dB.
2kHz= -9dB
4kHz= -15dB
8kHz= -21dB
16kHz= -27dB
32kHz= -33dB which is still a fairly high level. The cutoff slope is very gradual.